Silicon precision timing detectors for minimum ionizing particles FNAL-LDRD-2017-027 Artur Apresyan, Lindsey Gray All Experimenters' Meeting 08.28.2017
LDRD Project description • The aim is to develop over two years the technology that achieves – Time-tagging at σ t =20-30 psec for single MIPs, – Construct "large system" detector demonstrator, comprised of around 30-50 individual readout channels • Next generations of detectors for hadron colliders will face enormously challenging experimental conditions – At HL-LHC: 140-200 overlapping interactions per bunch crossing – FCC or similar 100 TeV collider: up to 2,000 pileup interactions • Extreme density of charged particles severely degrades event reconstruction: charged lepton eff., jets/MET resolution, etc.. 2 8/28/17 A. Apresyan | All Experimenters' Meeting
HL-LHC beam spread • Beamspot width in time is several hundred ps ( RMS = 200 ps ) – A detector with ~30 psec timing resolution could distinguish between interactions on the basis of timing 3 8/28/17 A. Apresyan | All Experimenters' Meeting
Event reconstruction at HL-LHC 78 pp collision LHC bunch crossing 1 ns interval 4 8/28/17 A. Apresyan | All Experimenters' Meeting
Event reconstruction at HL-LHC 78 pp collision LHC bunch crossing 1 ns interval 78 pp collision LHC bunch crossing 1 ns interval 0.5ns - 0.8ns - 0.11ns 0.11ns - 0.02ns 0.14ns 5 8/28/17 A. Apresyan | All Experimenters' Meeting
Muon reconstruction • Muon charged isolation efficiency in Z → μμ and ttbar (fake) events • Timing yields 10% improvement per muon 6 8/28/17 A. Apresyan | All Experimenters' Meeting
Precision Timing Detectors • MIP timing detector: cover up to | η |<3.0 to time stamp charged particles in the event: ~30 psec timing resolution • Timing detector in the endcap: – High granularity detectors needed in the forward region due to particle density – Radiation tolerance up to ~2x10 15 n/cm 2 to survive 3,000 fb -1 – Time resolution of ~30 pse for single MIPs • Barrel timing detector with SiPM+LYSOs 7 8/28/17 A. Apresyan | All Experimenters' Meeting
Timing layer in the endcap • Silicon sensor with specially doped thin region that produces high electric field à produces avalanche providing signal 15-30 gain – Large community: RD50 collaboration, several manufacturers (CNM, FBK, Hamamatsu) • Key Challenges our LDRD tries to tackle: – Achieve radiation tolerance up to 2x10 15 n eq /cm 2 at | η | = 3.0 for 3,000 fb − 1 – Develop a process to produce large area , uniform gain , high production yield LGAD sensors. 8 8/28/17 A. Apresyan | All Experimenters' Meeting
FNAL Readout Board • Developed a readout board for the characterization of LGADs – Final goal is to have tens of channels on one board: need to learn! – Similar boards developed previously by others for 1- or 2-channels – FNAL 4-ch board is cheaper, simpler, and is as good as the alternatives – FNAL board is now being used at UCSB, UCSC, KIT, CERN, and more are being prepared Hamamatsu 2x2 LGAD array3x3 mm 2 pixels FNAL readout board: 4 pixel HPK sensor Designed by S. Los 9 8/28/17 A. Apresyan | All Experimenters' Meeting
Sensor R&D and testing • Collaborative effort with CMS & ATLAS institutes: • Caltech, FNAL, Univ. of Kansas, Univ. of Torino, UC Santa Cruz • Close collaboration with Hamamatsu, CNM, FBK • Characterization of newest LGAD sensors: irradiated & unirradiated • Fermilab and LDRD was critical for the success of this campaign: FTBF, SiDet, support of technicians and engineers, readout board desing and production • Precision tracking detector available at FTBF: unparalleled precision of measurements 10 8/28/17 A. Apresyan | All Experimenters' Meeting
LGAD sensor uniformity • We observe a flat 100% efficiency across the whole sensor area. • A clear drop in efficiency is observed in the transition region between the two pixels 100% particle detection efficiency across the sensitive area Efficiency 1 LGAD Sensor: HPK 50D 0.9 Transition from one pixel to the other 0.8 Efficiency 11 12 13 14 15 16 17 18 19 20 1 0.9 LGAD Sensor: CNM W9HG11 0.8 11 12 13 14 15 16 17 18 19 20 x-coordinate [mm] 11 8/28/17 A. Apresyan | All Experimenters' Meeting
LGAD interpixel ”no-response area” • Thanks to the pixel telescope in the FTBF, for the first time we can look into the LGAD behavior between pixels – We measure the no-response width to be around 110 μ m on the HPK sensor. And around 70 μ m on CNM sensors. 12 8/28/17 A. Apresyan | All Experimenters' Meeting
LGAD sensor uniformity • Very uniform gain distribution across sensor surface – A flat response with a uniform signal size is observed over the whole sensor area 0.05 MIP MPV [V] LGAD Sensor: HPK 50D 0.045 0.04 Transition from one pixel to the other 0.035 0.06 LGAD Sensor: CNM W9HG11 MIP MPV [V] 10 11 12 13 14 15 16 17 18 19 0.05 0.04 0.03 10 11 12 13 14 15 16 17 18 19 x-coordinate [mm] 13 8/28/17 A. Apresyan | All Experimenters' Meeting
LGAD sensor uniformity • We observe a uniform time resolution around 40 ps across the whole surface area for HPK, and around 55 ps for CNM sensors. Time Resolution [ps] LGAD Sensor: HPK 50D 60 40 Time Resolution [ps] 11 12 13 14 15 16 17 18 19 100 LGAD Sensor: CNM W9HG11 50 11 12 13 14 15 16 17 18 19 x-coordinate [mm] 14 8/28/17 A. Apresyan | All Experimenters' Meeting
Comparison of HPK doping profiles • Dependence of the sensors’ characteristics on the doping concentrations were performed by comparing the 50 μ m HPK sensors of different gain splits. – Difference between doping concentrations of adjacent splits is about 4% Time resolution [ps] 80 HPK 50A-PIX 70 60 We observe a uniform time 50 40 resolution around 40 ps 30 Time resolution [ps] 11 12 13 14 15 16 17 18 80 HPK 50B-PIX across the whole sensor 70 60 area 50 40 30 Time resolution [ps] 80 11 12 13 14 15 16 17 18 Some difference between HPK 50C-PIX 70 adjacent sensors due to KU 60 50 board having variations on 40 30 between channels 11 12 13 14 15 16 17 18 Time resolution [ps] 60 HPK 50D-PIX 50 40 30 11 12 13 14 15 16 17 18 x-coordinate [mm] 15 8/28/17 A. Apresyan | All Experimenters' Meeting
Irradiated HPK sensors performance • Irradiation causes gain layer to fade – To preserve time resolution and gain, need to increase the operating bias voltage – Excellent uniformity of signal across the irradiated HPK sensor area • Time resolution slightly improves with the increase of the bias voltage, and shows a uniform distribution across the sensor, around 30 ps High uniformity of signal over irradiated sensor area LGAD Sensor: HPK 50D Time Resolution [ps] y-coordinate [mm] Mean Amplitude [mV] LGAD Sensor: HPK 50D 140 22.2 40 120 Bias Voltage: 600V 22 Bias Voltage: 635V 100 20 21.8 80 21.6 Time Resolution [ps] 13.5 14 14.5 15 15.5 16 16.5 LGAD Sensor: CNM W11LGA35 60 60 21.4 40 40 Bias Voltage: 400V 21.2 20 Bias Voltage: 420V 21 0 13.4 13.6 13.8 14 14.2 14.4 20 13.5 14 14.5 15 15.5 16 16.5 x-coordinate [mm] x-coordinate [mm] HPK 6 x 10 14 n.eq/cm 2 at 600V BV 16 8/28/17 A. Apresyan | All Experimenters' Meeting
Irradiated CNM sensors performance • Two distinct regions can be identified on the sensor based on the signal amplitude: – Different behavior under the aluminum metallization, and the region in the center • The highest bias voltage reached is − 420 V and the timing resolution is 30 ps for the metallized part and 40 ps for the non-metallized area. High uniformity of signal over irradiated sensor area LGAD Sensor: CNM W11LGA35 100 Time Resolution [ps] y-coordinate [mm] Mean Amplitude [mV] LGAD Sensor: HPK 50D 90 23 40 80 Bias Voltage: 600V 22.8 70 Bias Voltage: 635V 60 20 22.6 50 22.4 Time Resolution [ps] 13.5 14 14.5 15 15.5 16 16.5 LGAD Sensor: CNM W11LGA35 40 60 22.2 30 20 22 40 Bias Voltage: 400V 10 Bias Voltage: 420V 21.8 0 15.2 15.4 15.6 15.8 16 16.2 16.4 20 13.5 14 14.5 15 15.5 16 16.5 x-coordinate [mm] x-coordinate [mm] HPK 6 x 10 14 n.eq/cm 2 at 600V BV 17 8/28/17 A. Apresyan | All Experimenters' Meeting
Summary • Excellent start to the LDRD program – Extremely successful test beam campaign, many first measurements – 1 st paper on sensor performance is to be submitted by end of August. 2 nd paper on board performance to follow soon • Measurement presented (will be) at several conferences – RD50 collaboration meeting, AWLC2017, Hiroshima symposium – Collaborations with various institutes established • We have contacted the manufacturers (CNM, HPK, FBK) to proceed to the next stage of the R&D targeting larger sensor arrays – Expect next batch of production within 6-8 months 18 8/28/17 A. Apresyan | All Experimenters' Meeting
Backup 19 8/28/17 A. Apresyan | All Experimenters' Meeting
Vertex reconstruction with timing 200 PU 140 PU LHC • ∼ 5x reduction in effective pileup in terms of charge multiplicity 20 8/28/17 A. Apresyan | All Experimenters' Meeting
Recommend
More recommend
